43 results about "Subdivision algorithms" patented technology

A method for the computation of a subdivision surface from a base mesh of coarse faces which requires no dynamic allocations is performed as follows. First, a static data structure is allocated. The static data structure includes a hierarchy of data arrays, where each of the data arrays is for storing control mesh data for a single coarse face at a respective subdivision level. The size of each of the data arrays is determined by the respective subdivision level. A subdivision algorithm is then applied to each of the base mesh coarse faces in turn. During the subdivision of a single coarse face, the resulting data for each subdivision level is stored in the respective data array. The static data structure may store further data for each subdivision level respectively, such as a tag for each of the sub-faces at the given level. The subdivision may be applied adaptively, by applying level-of-detail control information at intermediate subdivision levels.

The invention provides a method for processing medical images, comprising the following steps: A. c preprocessing two-dimensional images, and partitioning and extracting the preprocessed two-dimensional images; B. establishing surface mesh and body mesh moulds, adopting a subdivision algorithm for processing, and establishing a visualization platform; and C. adopting a light projection algorithm based on VTK for processing the two-dimensional images, and obtaining three-dimensional visualization images. When the method for processing medical images provided by the invention is adopted, fast batch processing can be carried out on the medical two-dimensional images, and three-dimensional visualization of the two-dimensional images can be realized.

A method for predicting user loss of an insurance e-commerce platform comprises the following steps: step 1, collecting original data of a website client, and storing the original data in a data warehouse according to a user rule; Step 2, performing cleaning, integration and protocol preprocessing on the original data, and further extracting variables for measuring the user value from the integrated data set for user subdivision; Step 3, selecting a reasonable observation variable and a proper user subdivision algorithm to carry out user classification; 4, selecting corresponding variables influencing user loss for different user groups, and performing user loss probability prediction respectively; 5, different prediction algorithms are selected for different types of users respectively, the model effect is evaluated through indexes such as the accuracy rate and the recall rate, and when the model effect is optimal, the final loss probability of the different types of users is output;And step 6, performing classified management on different types of lost user groups, performing group feature description respectively, and providing data reference for revocation strategy design so as to realize refined marketing and perform subsequent marketing effect analysis later.

The invention discloses a digital image measurement sub-pixel edge processing method which carries out a bulk processing on the image edge by adopting a gradient model, takes a gradient mean as the threshold of the binaryzation control, realizes a search of binaryzation images in the biggest connected region on the basis of a flood filling principle and searches consecutive points of which a segment of gradient values exceeds the gradient threshold at all location points in four direction; the direction which has the smallest number of consecutive points is the direction of normal; a gradient peak value along the direction of normal is calculated on base of a spline interpolation subdivision algorithm; edge points contained on a fitting curve is marked from the collected target rough edges; after multiple rounds of iterative computations, if all points on the collected target rough edges are marked, the precise sub-pixel edge computation is finished. The processing method of the invention bases on traditional gradient edge extraction, flood filling algorithm and spline interpolation subdivision algorithm; the processing course is realized through a crude image processing and a refined image processing; the target edge is searched on purpose; the computational efficiency is high and the algorithm course takes the requirement of the noise suppression into consideration and increases the immunity of the algorithm.

The invention discloses a method for defogging single images based on Bayer formats rapidly. The method comprises using a quad-tree subdivision algorithm to estimate atmospheric light; calculating atmospheric transmission images based on dark channel prior models, and using Guided Filter to replace Soft Matting to optimize the atmospheric transmission images to further calculate fogless Bayer images; and using a demosaicing algorithm to obtain clear fogless images. According to the method, images of Bayer formats are defogged directly, original data are reduced by 2 / 3 than those through other methods, and the calculation speed is increased greatly on the premise of guaranteeing defogging effects.

The invention discloses a novel moire fringe high-power subdivision method based on less spatial points, which is applicable to wide-range displacement quick and precise measurement, positioning and control. The position of a grating is determined through the four-way spatial signals output by a grating sensor at a certain moment, and the total displacement is obtained by accumulating the position differences of the two neighboring moments; and the high-resolution high-power subdivision is obtained by adopting the high-precision high-speed data collecting system and the space less point subdivision algorithm. The method fully utilizes the cyclicity and sine property of the moire fringe, adopts the FPGA (Field Programmable Gate Array) to realize algorithm, has the characteristics of less calculation, high speed and high subdividing number, and can solve the contradictory problems in grating wide-range precise displacement measurement and control of high resolution, high speed and wide range.

The invention discloses a point cloud mesh refinement system and a point cloud mesh refinement method, which are applied to a computer device. The method comprises steps: a point cloud for a target product is acquired, and triangulation is carried out on the point cloud for the target product; triangular meshes in need of refinement are selected, and a needed iteration number N is inputted according to the precision requirement for the mesh refinement; according to the shape and the position of each triangular mesh, a mesh subdivision method is used for subdividing each triangular mesh into different types; a loop triangular mesh subdivision algorithm is used for carrying out N times of refinement processing on the triangular meshes of different types, and triangular meshes after refinement are obtained; point cloud smoothing processing is carried out on the triangular meshes after refinement, a three-dimensional mesh model for the target product is obtained, and the three-dimensional mesh model for the target product is displayed on a display device. When the system and the method of the invention are executed, quick and effective refinement can be carried out on the point cloud mesh of the target product, the model features of the target product can be kept, and a refiner and smoother three-dimensional mesh model for the target product can be acquired.

A system, method, and computer program product are provided for tessellation using shaders. New graphics pipeline stages implemented by shaders are introduced, including an inner ring shader, an outer edge shader, and topologic shader, which work together with a domain shader and geometry shader to provide tessellated points and primitives. A hull shader is modified to compute values used by the new shaders to perform tessellation algorithms. This approach provides parallelism and customizability to the presently static tessellation engine implementation.

The invention belongs to the field of computer graphics and data visualization and relates to a seismic data three-dimensional display method based on surface subdivision. The seismic data three-dimensional display method is realized by utilizing a Directx11 programming tool and based on a third-order Bessel surface subdivision algorithm. For prepared vertex buffer, a GPU is set to enable control points to be assembled in a surface patch manner, so that index buffer is not needed, and the problem of grooves and ridges generated in the process of determining the index buffer is prevented; and besides, through the application of the surface subdivision technology based on GPU programming, more control points are generated, and display of a grid is allowed to be finer and smoother and have better real-time performance.

The invention discloses an acceleration method of a ray tracing algorithm in home decoration design software. The optimization of the ray tracing algorithm and the hardware acceleration of a GPU (Graphics Processing Unit) are combined, the optimized ray tracing algorithm is realized through CUDA (Compute Unified Device Architecture) programming, the ray tracking algorithm based on CUDA acceleration is realized on the high-grade GPU which supports recursion and the low-grade GPU which does not support the recursion, and the software and hardware combination of the optimization of the algorithm and hardware acceleration of the GPU is used for accelerating the rendering of the ray tracking algorithm. The acceleration method fully performs the advantages of a bounding box algorithm and a space subdivision algorithm, solves the blindness problem of ray and bounding box traversal, is simple, is obvious in acceleration effect, is simple in implementation, avoids a phenomenon that a great quantity of data is transmitted between a CPU (Central Processing Unit) and the GPU, fully utilizes the own memory of the GPU, and provides a superexcellent implementation scheme for the algorithms including the ray tracking algorithm and the like which operate on the CUDA and need recursion.

The invention relates to the technical field of motors, and provides a stepping motor driver and a random subdivision algorithm therefor. The algorithm comprises the following steps: setting a pulse equivalent as a constant, obtaining a phase range according to a subdivision value, wherein the phase range is n times of the pulse equivalent, wherein n is an integer; receiving a pulse signal, and obtaining a relative pulse speed according to the pulse signal and the pulse equivalent; obtaining a current phase value according to the relative pulse speed, wherein the current phase value is m times of the subdivision value, wherein m is an integer; converting the current phase value into a corresponding angle in a phase range of a sine table according to the subdivision value; reading the data of the sine table according to the angle, and outputting a phase current. The method achieves a purpose that the variables will automatically change after a user randomly selects the subdivision, thereby avoiding the later manual operation.

The invention discloses a voltage flicker detection algorithm based on improved HHT (Hilbert-Huang Transform). The voltage flicker detection algorithm comprises the following steps: 1) extracting voltage envelope signals from collected discrete voltage signals by using a Hilbert transform method; 2) in an empirical mode decomposition (EMD) 'screening' step, 'splitting' a new control point for three times of spline interpolation fitting of an envelope line by using a four-point interpolation subdivision algorithm, and optimizing the new control point by the method; and 3) decomposing the voltage envelope signals by using improved EMD so as to obtain a group of intrinsic mode functions (IMF), and extracting flicker amplitude values and flicker frequency parameters from the group of decomposed IMF by using the Hilbert transform method respectively. The voltage flicker detection algorithm disclosed by the invention effectively overcomes the problem that parameter detection errors are increased due to unsatisfactory IMF generated by EMD, thereby obtaining a parameter detection result with higher precision.

The invention provides a processing circuit for two-dimensional plane displacement measurement, and relates to the field of photoelectric displacement precision measurement. An image data receiving module receives image pixel data transmitted by external equipment through an image data receiving connector, and transmits the image pixel data to a code recognition module and a subdivision operationmodule respectively, the code recognition module is used for recognizing and decoding the image pixel data, the subdivision operation module is used for carrying out two-dimensional subdivision operation on the image pixel data, a data integration module is used for receiving the image pixel data recognized, decoded and subdivided by the code recognition module and the subdivision operation moduleand then fusing the data, and a data output module sends the fused data to receiving equipment through a data output connector. Through cooperation with the external equipment, two-dimensional codinginformation of a current measurement position can be accurately identified, high-resolution displacement subdivision is realized based on a sub-pixel-level subdivision algorithm, and high-resolutiontwo-dimensional displacement measurement is further realized.

The invention relates to the technical field of machine manufacturing and the technical field of image measurement in mechanical engineering, and in particular relates to the field of three-dimensional metallographic testing. A three-dimensional modelling measurement method of a microscopic multi-phase structure based on a section profile sequence comprises the steps of: making a microscopic multi-phase structure material specimen into a specimen for metallographic testing, trying to grind through a first grinding wheel, a second grinding wheel and a third grinding wheel, simultaneously, acquiring an upper surface image of the specimen through a first microscopic lens, a second microscopic lens and a third microscopic lens, determining profile data of each phase boundary in a faultage image by using edge detection and sub-pixel subdivision algorithms, performing three-dimensional model reconstruction of the closed boundary of each phase by using eight-field joint mark tracking and NURBS freedom curve reconstruction methods, and obtaining a three-dimensional entity model of each phase through Boolean operation.

The invention discloses a simulation modeling method and system based on a scene tree, and the method comprises the steps of obtaining simulation data according to a simulation target, initializing a data structure of the simulation data, and obtaining object data and scene data included in the simulation data; constructing a scene tree according to an AABB bounding box of the environment data and the object data, performing spatial hierarchy subdivision on the constructed scene tree based on a scene tree spatial subdivision algorithm, and establishing a scene tree spatial index; establishing topological connection between the object nodes according to the scene tree spatial index and other object nodes connected with one object node in the scene tree; and after the topological connection relationship is established, writing classification information and attribute information of each object node, and establishing a simulation model. The development complexity of the object nodes can be effectively controlled while a consistent access method for the object nodes is achieved, data are stored in a centralized mode through a simulation modeling method of the scene tree, and management and visual development of the data are facilitated.

The invention discloses a spot welding robot operation space smooth path planning method for a curved surface workpiece. In the method, a spot welding robot kinematic model module, a motion constraint condition module, an inter-welding-spot shortest smooth obstacle avoidance path planning module and an optimal welding spot welding sequence planning module are included. The method is characterized by modeling through a three-dimensional grid method curved surface workpiece profile and welding spot distribution, adopting an improved A-star algorithm and a uniform B spline curve subdivision algorithm for smooth processing to generate an inter-welding-spot shortest smooth obstacle avoidance path, and applying a multi-target elite simulated annealing genetic algorithm to obtain an optimal welding sequence; and according to the collision-free motion constraint conditions of an electrode holder coordinate system and a welding spot coordinate system and the safe distance constraint conditions of the electrode holder coordinate system and a curved surface workpiece profile, solving through inverse kinematics to obtain a joint space path corresponding to the current welding path. The method has an application reference value in the actual industry, the planning and debugging time of an engineer can be shortened, and the working efficiency of a robot can also be improved.

The invention relates to a selection behavior preference subdivision algorithm. According to the algorithm, model parameters are estimated on the basis of selection behavior data and a discrete selection model so as to serve as representations of selection behavior preferences. The algorithm comprises the following steps of: decomposing the data into a plurality of sub-samples which are solvable to the lowest limit by using a Monte Carlo algorithm, wherein basis of the decomposition is a similarity between the model parameters, namely, comprehensively considering a relative difference betweentwo model parameters and credibility of the difference; classifying all the sub-samples step by step which are solvable to the lowest limit by taking the similarity as a distance index by using a hierarchical clustering algorithm; and finally determining a proper quantity of models according to a Bayesian information standard so as to obtain subdivided selection behavior preference types. According to the algorithm provided by the invention, heterogeneity in selection behavior data is mined; and compared with basic discrete selection model algorithm, the algorithm is capable of obtaining moreaccurate selection behavior preferences and models of different types of persons or groups, and laying a foundation for obtaining more accurate and pointed behavior prediction results for related applications.

A system, method, and computer program product are provided for tessellation using shaders. New graphics pipeline stages implemented by shaders are introduced, including an inner ring shader, an outer edge shader, and topologic shader, which work together with a domain shader and geometry shader to provide tessellated points and primitives. A hull shader is modified to compute values used by the new shaders to perform tessellation algorithms. This approach provides parallelism and customizability to the presently static tessellation engine implementation.

An earth and rockfill dam complex borrow pit excavation volume calculation method based on a grid subdivision algorithm comprises the following steps that three-dimensional point cloud data of a borrow pit excavation face are scanned, registration of point clouds of different scanning batches is achieved through a point cloud registration algorithm, so that point clouds which are originally and irregularly distributed on the excavation face are obtained; the obtained original irregularly distributed point cloud is processed through a Doo-Sabin subdivision algorithm, and the point cloud forms a relatively regular excavation surface; then, Kriging interpolation calculation is performed to obtain a spatial point cloud participating in excavation volume calculation; a grid subdivision algorithm is compiled through MATLAB, and calculation of the excavation volume is achieved. According to the earth and rockfill dam complex borrow pit excavation volume calculation method based on the grid subdivision algorithm, the borrow pit excavation volume can be rapidly estimated, and the method has the advantages of being small in workload, relatively safe in measurement and high in efficiency.

The invention relates to the field of hydropower dispatching operation, in particular to an automatic avoiding method for a high-water-head irregular multi-limiting area of a main stream cascade hydropower station group. According to the technical scheme, on the basis that mathematical definition is carried out on constraint of a unit limiting area, automatic analysis technology subdivision is carried out on an irregular limiting area through a Hertel-Mehlhorn-based convex subdivision algorithm, and then modeling is carried out on a subdivision result through a convex optimization theory and adisjunction planning theory. According to the method, irregular multi-limit areas can be automatically avoided, and the effects of automatic modeling and solving are achieved. The method has important significance for responding to power grid differentiation and automatic dispatching application in the future and responding to the new situation of frequent change of hydropower output under the market condition.

The invention discloses a method for processing an obj file on the basis of a vtk (Visualization Toolkit). The method comprises the following steps that: S1: reading a surface patch in an obj file, and constructing a reverse index table; S2: searching a shared vertex; S3: generating a new vertex, updating and constructing surface patch id (Identification) if the shared vertex is found, if no shared vertexes are found, generating a new vertex which has the same position, and jumping to S4; S4: increasing the surface patch, generating the new vertex, adding the surface patch for connecting the new vertex with the old vertex for continuity between the new vertex and the old vertex, and adding the new vertex into a vertex set; S5: subsequent vertexes and subsequent surface patches: if the subsequent vertex of the surface patch is not processed, jumping to S2, and if the subsequent surface patch of the model is not processed, jumping to S1; S6: triangular surface patch processing: splitting all surface patches into triangular surface patches; and S7: hook face smoothness: applying a loop hook surface subdivision algorithm to smoothen the model.

The invention discloses a geologic body variable attribute interpolation method and system. The geologic body variable attribute interpolation method comprises the steps of loading interpreted geologic body result data in a set work area range; constructing a surface grid model of the geologic body according to the control point data of the geologic body; taking the block geometric surface triangular mesh data as a limiting condition, carrying out regularization processing on the limiting condition to obtain regular limiting points, limiting line segments and limiting surfaces, and obtaining aregularized surface mesh model; subdividing the regularized surface mesh model by adopting a three-dimensional limited DELANAY subdivision algorithm to obtain a tetrahedral mesh with topological consistency; and performing attribute interpolation on the tetrahedral mesh by using a gradient-based variable attribute interpolation algorithm to obtain an interpolation result. According to the geologic body variable attribute interpolation method and system, variable attribute interpolation calculation of a complex geologic body is achieved, modeling efficiency is high, practicability is high, thepurpose of quick attribute modeling is achieved, and the geologic body variable attribute interpolation method and system are suitable for parallel calculation of a big data body attribute model.

The invention discloses an automatic portrait slimming method based on 3D modeling. The automatic portrait slimming method comprises the following steps: acquiring position coordinates, genders and weight grades of all human bodies in a picture by using a self-training human body detection algorithm; for each to-be-processed human body area, obtaining skeleton key points of a figure main body, andremoving a background; matching the human body area with the background removed with the key point coordinates, sending into a 3D modeling algorithm to obtain a 3D model of the human body; obtaininghuman body shape parameters output by the 3D model through PCA principal component analysis, and achieving slimming operation on the 3D model; projecting the 3D model completing slimming onto a 2D image; and obtaining a triangular mesh by using a Delaunay triangle subdivision algorithm to obtain a final slimming result. According to the invention, the step of manually liquefying and slimming by using software is greatly simplified, and the difficulty that a mainstream automatic slimming function cannot be used for carrying out fine-grained adjustment on each part of the body is overcome.

The invention discloses a parameterized false tooth coronal internal surface generation method. The parameterized false tooth coronal internal surface generation method comprises the steps of S1, obtaining a prepared tissue surface of a false tooth coronal, and obtaining preset generation parameters of a false tooth coronal internal surface at the same time; S2, configuring a reference cross section of the prepared tissue surface; S3, generating multiple regression line rings from the prepared tissue surface according to the reference cross section, and meanwhile, conducting region dividing onthe prepared tissue surface according to the preset generation parameters; S4, conducting non-equidistant bias on the divided regression line rings in different areas according to the preset generation parameters; S5, matching the biased regression line rings; S6, adopting a skin subdivision algorithm to generate the false tooth coronal internal surface from the matched regression line rings. Bymeans of the parameterized false tooth coronal internal surface generation method, the prepared tissue surface can be subjected to region segmentation, then different areas of the top, the side surfaces and the bottom of the prepared tissue surface are processed, and finally the false tooth coronal internal surface is generated. Optimal fitting can be achieved, and the parameterized false tooth coronal internal surface generation method is small in calculation amount and high in precision, and can be widely applied to the false tooth repair industry.

The invention relates to a selection behavior preference subdivision algorithm. According to the algorithm, model parameters are estimated on the basis of selection behavior data and a discrete selection model so as to serve as representations of selection behavior preferences. The algorithm comprises the following steps of: decomposing the data into a plurality of sub-samples which are solvable to the lowest limit by using a Monte Carlo algorithm, wherein basis of the decomposition is a similarity between the model parameters, namely, comprehensively considering a relative difference betweentwo model parameters and credibility of the difference; classifying all the sub-samples step by step which are solvable to the lowest limit by taking the similarity as a distance index by using a hierarchical clustering algorithm; and finally determining a proper quantity of models according to a Bayesian information standard so as to obtain subdivided selection behavior preference types. According to the algorithm provided by the invention, heterogeneity in selection behavior data is mined; and compared with basic discrete selection model algorithm, the algorithm is capable of obtaining moreaccurate selection behavior preferences and models of different types of persons or groups, and laying a foundation for obtaining more accurate and pointed behavior prediction results for related applications.

The invention relates to a method and a device for precisely measuring ultrasonic wave transmission time. The method adopts two ultrasonic wave transducers A and B, a hardware circuit and a software subdivision algorithm. The hardware circuit mainly comprises an ultrasonic wave transducer drive circuit, an ultrasonic wave return signal filter circuit, an amplification circuit and a signal processing circuit. The signal processing circuit comprises an analog-to-digital (A / D) converter, an FPGA and a CPU. The CUP control the FPGA to start the ultrasonic wave transducer drive circuit to drive the transducer A to send an ultrasonic wave signal; the filter circuit filters and amplifies an ultrasonic wave return signal received by the ultrasonic wave transducer B; the A / D converter samples the wave return signal and stores the sampled data in a storage region constructed in the FPGA; and after the sampling is completed, the CPU reads the sampled data from the FPGA and accurately calculates the transmission time of the ultrasonic wave between the two transducers A and B by adopting the software subdivision algorithm. As the FPGA-based hardware circuit and the special software subdivisionalgorithm are adopted, the method and the device can realize the measurement of the ultrasonic wave transmission time with nanosecond precision and ensure good real-time.

The invention discloses a digital image measurement sub-pixel edge processing method which carries out a bulk processing on the image edge by adopting a gradient model, takes a gradient mean as the threshold of the binaryzation control, realizes a search of binaryzation images in the biggest connected region on the basis of a flood filling principle and searches consecutive points of which a segment of gradient values exceeds the gradient threshold at all location points in four direction; the direction which has the smallest number of consecutive points is the direction of normal; a gradientpeak value along the direction of normal is calculated on base of a spline interpolation subdivision algorithm; edge points contained on a fitting curve is marked from the collected target rough edges; after multiple rounds of iterative computations, if all points on the collected target rough edges are marked, the precise sub-pixel edge computation is finished. The processing method of the invention bases on traditional gradient edge extraction, flood filling algorithm and spline interpolation subdivision algorithm; the processing course is realized through a crude image processing and a refined image processing; the target edge is searched on purpose; the computational efficiency is high and the algorithm course takes the requirement of the noise suppression into consideration and increases the immunity of the algorithm.

The invention provides a curved surface grid generation method and device, equipment and a storage medium. The method comprises the following steps: constructing a triangular manifold grid on a curved surface; determining a corresponding subdivision node according to each grid edge of the triangular grid; the subdivision base point is located on the curved surface; connecting subdivision nodes corresponding to the grid edges forming the same triangle in the triangular grid, and obtaining subdivision grid edges; and constructing the curved surface grid by adopting the subdivided grid edges. According to the curved surface grid generation method and device, the electronic equipment, the storage medium and the computer program product, the characteristics of the extension direction of the curved surface do not need to be considered during use, the grid can be directly generated based on the nodes on the curved surface, and rich grid patterns can be generated.

The invention relates to the field of hydropower dispatching and operation, in particular to a method for automatically avoiding high-head, irregular and multi-restricted areas of cascade hydropower station groups in main streams. The technical solution is: based on the mathematical definition of the constraints of the restricted area of ​​the unit, the irregular restricted area is segmented by automatic analysis technology based on the Hertel-Mehlhorn convex subdivision algorithm, and then the convex optimization theory is used for the segmented results Modeled with disjunctive programming theory. This method can automatically avoid irregular and multi-restricted areas, achieving the effect of automatic modeling and solving. This method is of great significance for responding to grid differentiation, automatic dispatching applications, and responding to the new situation of frequent changes in hydropower output under market conditions in the future.

The invention discloses a method for processing an obj file on the basis of a vtk (Visualization Toolkit). The method comprises the following steps that: S1: reading a surface patch in an obj file, and constructing a reverse index table; S2: searching a shared vertex; S3: generating a new vertex, updating and constructing surface patch id (Identification) if the shared vertex is found, if no shared vertexes are found, generating a new vertex which has the same position, and jumping to S4; S4: increasing the surface patch, generating the new vertex, adding the surface patch for connecting the new vertex with the old vertex for continuity between the new vertex and the old vertex, and adding the new vertex into a vertex set; S5: subsequent vertexes and subsequent surface patches: if the subsequent vertex of the surface patch is not processed, jumping to S2, and if the subsequent surface patch of the model is not processed, jumping to S1; S6: triangular surface patch processing: splitting all surface patches into triangular surface patches; and S7: hook face smoothness: applying a loop hook surface subdivision algorithm to smoothen the model.